KOBERTs) ANALYSIS OF SONGS 401 



pitched intervals. Although these songs are too few to make possible 

 the positive denial or assertion of any such practice, experience with 

 the purely vocal music of other aboriginal peoples of comparatively 

 low culture would decidedly favor the assumption that they do not. 



Apparently, regular metric (and also rhythmic) patterns are lilve- 

 wise not formed but forming. By the word metric pattern is meant 

 a succession of metric groups (measures) either all the same size or, 

 if of different lengths, reappearing m regular order regardless of 

 melodic, rhythmic or word accompaniment; such, for instance, as a 

 regular succession of 2, 3, 4, 2, 3, 4 meters. By rhythmic pattern is 

 meant recurrence, withm natural division of the melody, of a group 

 of notes of contrasting lengths, in the same order as previously set 

 forth; a recurrence not incidental to direct repetition of nielody and 

 words, but which reappears despite changes in either. Were such a 

 rhythmic persistency discovered, it would be fair evidence of the 

 rhythmic feature being enjoyed and used for its own sake and not 

 by chance. 



Now and then songs are encountered in which the meter is almost 

 regular, but in many it gives way in places to the demands of the 

 text, and beats are lost or inserted which temporarily disturb the 

 swing of the movement. It is interestmg to observe occasionally, 

 however, that metric shifts are apparently disturbing to the singer, 

 for a beat lost in one measure is now and then added to the next, 

 and vice versa. Regularity of meter is more apt to occur in songs 

 which are generally accompanied with regularity of bodily move- 

 ment, such as marchuig songs, rather than with incantations, for 

 instance. 



I have retained the use of measure bars in writing the songs, as 

 indicating the return of tiie pi-imarj^ accent. They are a great help 

 in reading the music and much to be preferred to accent marks above 

 a long series of otlierwise undivided notes, especially as such marks 

 are sometimes needed for other purposes than for indicating pri- 

 mary accents. Constantly changing time signatures, including those 

 which contain fractional beats, are therefore self-explanatorj'. It is 

 troublesome to insert them at each change, and to omit them is an 

 error easily committed, but they seem preferable in reading to the 

 omission of measure bars. 



In this collection, out of several renditions of the same song, one 

 usually showed a greater tendency to definiteness and regularity 

 than the others. Although these considerations might be argued as 

 rather arbitraiy standards by wliich to decide on the best rendition, 

 especially according to Picuris ways of thinking, since it was not 

 possible to question the singer as to which rendition he preferred, 

 I have used them in making my own selection of the rendition which 

 was to be placed with the story. There are 11 story songs. The 



